Well there are a couole of points to note. First  HV earthing is not simple - now in the UK,  in a buried cable the HV earth usually comes with the supply and has a low impedance back to the transformer,  where it is quite common to have a limiting impedance, say 1 ohm so that max fault current on  a short run is about 6000A from an 11kV supply (each phase 6k5 to ground and 11k phase to phase) But overhead is normally 11kV phase wires only no earth or neutral, so the HV earth electrode then carries the full can to operate expulsion fuses or what is in effect the worlds largest RCD at the supply end. (though  it is done with separate current transformers and special breakers)

At the end of a long run (upto perhaps 10km on an 11kV line) the potential fault current is a lot lower - think hundreds of amps.

During the time the HV ADS is operating the region around the overhead fed  11kV /400V transformer is distinctly hot - not the full 6000 volts of course but equally well over the few hundred the LV side would be able to shrug off. The solution to such a 'hot site' is to have separated HV and LV earth - the LV earth may be running down the next pole along to the one with the transformer on. Of course a town site with underground HV and a probbaly a pretty substantual earth mat is a 'cold site' where rise of earth is not expected to be much, Then HV and LV earth can share metalwork.

The field around any electrode of any shape dies off as the distance increases, and from far enough away all earth structures are point like - but this is like saying if they could grow long enough hair all animals would be the same size and spherical - great on the blackboard but not useful  in theory, as you can almost never be 'far enough' away.

Equally most of the voltage drop is in the near field region and that is where we need to worry about step voltage etc.

(so put 230V into a typical 4ft rod, and (ignoring the electricity bill and the moral wisdom..) if you plot contours of declining voltage, once you get more than a rod length or so away, the less than 50V are left. (unless you have odd ground with buried pipes or something).

The earth mat under the substation may have dimensions of 10m or so, and the near field region is comparable, but very dependent on the detail of the shapes.

The trick is to keep these near field regions from overlapping if the electrodes are to be independant.

M.

Hi M, thank you for the reply. It is this near field I'm interested in know, this way I can ensure my TT rod is not overlapping. BS7430 shows a graph of potential fall, but is still at 10% at 10m. Looking at the graph does not seem to show getting anywhere close to 0%. So as you say you can never be far "enough away" We have had some feed back for the DNO, but they do not record the position of there HV earth.

It would ne nice to be able to say "Earth rod to be no closer than x m", so far finding x is the problem.

Forgot to mention rural 'pole pg' TT is not entirely risk free either, and the relatively high DNO electrode resistance can caus things to go a bit wild, especially when as is common, farms are augmented by non farm buildings like houses and chalets share, and folk 'improve' matters by PME-ing the houses - consider the fault described in this thread where the driving fault and the people being shocked are not even the same installations !!

It is rare but not unknown - another one

The DNO LV earth on a separated system can be as high as 20 ohms measured to the plate at the end of the universe, but single figure to ten ohms is more common.

If everything is TT, and the RCDs are all double pole and regularly tested,. it is a very safe system but mixed systems, or badly maintained ones, are another matter.

M.

While the DNO likely don't know where their electrode is, they should be able to tell you whether it's a hot or cold substation, because if it's hot it should be on the hot site register. Hot/cold relates to the EPR, which doesn't give you all the answers but does give you a hint as to scale of the issue, if there is one (and there may also be explicit exclusions along the lines of "Don't install anything earthed within this area")

Each DNO will have its own policy, for internal use, on combined and separate earthing systems; for example WPD/NGED TP21D/3. Of course that only applies to their network wheras a new TT rod will not be within scope. Any bare connection to earth within the zone may well be verboten under the terms of the connection agreement (not least to protect everyone else on the supply from transferred potential).

Note that if you are importing an LV supply from further away back into the hot zone, you might actually also be in need of an isolation transformer, since not only will E be referenced to some other bit of soil, but by definiton also the transformer star point and hence L123N, so you could suddenly see L or N swing high with respect to your local earth. This happens, for example, on DNO substations with an auxiliary supply fed from the LV network, but also on some commercial sites too.

this voltage will be present at the DNO head cut out. Which will effectively cause a potential between the DNO earth and local earth.

I would consider a TT earth to mitigate the above.

Pros and cons. TT keeps your earthing system close to 0V if you like, but L & N will still be referenced to the substation's (transformer) earth - so while there's less of a p.d. between PE and true earth in your installation, there can be a very big difference between L/N and PE - so the risk of damage to equipment can be significantly higher (especially stuff that references PE for functional reasons - e.g. things with EMI filters or some Data/telecoms equipment).

On a TN system L/N/PE would all go up and down together as it were - and as long as you're within an equipotential zone, the shock risk isn't any higher.

    - Andy.

The Problem is the near field is very situation dependent, so the text book authors and lecturers cough a bit and maybe consider easy cases like a single round rod in uniform semi-infinite soil, and then move on. But, if you know the largest dimension of your earthing system, then  you can assume it will always do better than some worst case structure of that size, probably  a huge rod of the diameter equal to the site extent, and rather like estimating an upper limit for the weight of a horse by pretending it is a sphere filled with cement you have a safe, if dog rough, rule of thumb, its just a very loose fit to reality. It is where the theoreticians' and experimentalists' views diverge..

The thing like a spherical horse approximation for earths, says you need to be a few 'maximum dimension equivalent distances away to be affected at the less than 10% level.

Now, who says 10% volage rise is never goodr enough ? You allow all earthed metal in a PME  LV system to rise to something like half mains voltage for the ADS trip time. (400ms for safety of life, 5 seconds for survival of equipment)

HV deserves a similar logic.- if the HV ADS is prompt - and what that really means varies a bit, depending on what may be affectrd then the 'safe' ground bounce can be quite high - some hundreds of volts - at some hot sites well over 600v - but of course only for a suitably short and very well defined period.

I'm being a bit arm wavy, but the whole topic is shrouded in the fog of what constitutes 'acceptable' levels of risk, and what those levels are, varies a bit.

Mike